MCP Time Tracking for AI Assistants

Connect Claude, Gemini, ChatGPT, and more to your TrackingTime workspace

The TrackingTime MCP Server gives your AI assistant direct access to your workspace. Log hours, list projects and tasks, query team reports, and create entries using natural language. The server is fully hosted, requires no installation, and is free for all TrackingTime plans.

Get Started

14-day PRO trial included

google icon Sign in with Google

About claude

MCP (Model Context Protocol) is an open standard that lets AI assistants connect to external services and act on real data. The TrackingTime MCP Server implements this protocol so your AI tool can read and write directly to your TrackingTime workspace. Every time entry, project, task, and report is accessible through natural language. The server runs on TrackingTime's infrastructure and is available to all users at no extra cost.

https://mcp.trackingtime.co/mcp

What is the TrackingTime MCP Server?

The TrackingTime MCP Server is a hosted service at https://mcp.trackingtime.co/mcp that connects AI assistants to your TrackingTime workspace using the Model Context Protocol. It supports HTTP and stdio transport and lets tools like Claude, Gemini, or Cursor log time, manage projects, query reports, and handle workspace data through natural language.

Which AI tools are compatible with the TrackingTime MCP Server?

Any MCP-compatible client can connect to the TrackingTime MCP Server. Confirmed integrations include Claude Desktop, Claude Code, Gemini CLI, ChatGPT (Pro and Plus with Developer Mode), Cursor, Windsurf, VS Code with GitHub Copilot, and Cline. The server supports both HTTP and stdio transport, so any tool that implements the MCP standard can connect.

Do I need to install anything to use the TrackingTime MCP Server?

There is nothing to install on TrackingTime's side. The server is fully hosted and always available at https://mcp.trackingtime.co/mcp. For clients using stdio transport, Node.js 18 or later is required to run the mcp-remote bridge via npx.

Is the TrackingTime MCP Server free?

Yes. The MCP Server is available to all TrackingTime users on any plan. You need a TrackingTime account and an App Password generated from your workspace settings to authenticate.

What can I do with TrackingTime through an AI assistant?

Once your AI assistant is connected via MCP, you can log time entries with start and end times, list active projects and tasks, query your team's hours for any date range, create projects, tasks, and customers, start and stop timers, and pull data for reports. All of this is accessible through natural language inside your AI tool.

CONNECT WITH CLAUDE

Start Tracking Time Directly from Claude Desktop or Code

The TrackingTime MCP Server connects to Claude through a simple configuration file. Setup takes less than five minutes and works for both Claude Desktop and Claude Code.

01

Generate an App Password

Log in to your TrackingTime workspace, go to Settings, and open Apps and Integrations. Create a new App Password and copy it. This is the credential Claude will use to authenticate requests to your workspace.

App settings
02

Add TrackingTime to Claude

Open Claude Desktop, go to Settings, then Developer, and click Edit Configuration.
Add the TrackingTime MCP server entry to your configuration file, including the server URL and your App Password.

MCP configuration
03

Restart and verify connection

Save the file and restart Claude Desktop. Go back to Settings, then Developer, and confirm that the TrackingTime server appears with an active status. You can now ask Claude to log time, list your projects, or query your team's hours.

Go to Claude

CONNECT WITH GEMINI

Add the TrackingTime MCP Server to Gemini CLI

Gemini CLI supports remote MCP servers via HTTP. You can add TrackingTime with a single terminal command or by editing your settings file directly. Both options work for global and project-level configurations.

01

Generate an App Password

Log in to your TrackingTime workspace, go to Settings, and open Apps and Integrations. Create a new App Password and copy it. This is the credential Gemini CLI will use to authenticate requests to your workspace.

App settings
02

Add TrackingTime to Gemini CLI

You can add the TrackingTime MCP server by running a single command from your terminal, or by adding an entry to your Gemini CLI settings file manually.
The server connects via HTTP, so no local bridge is required.

MCP configuration
03

Verify the connection

Inside Gemini CLI, run /mcp to check the connection status. TrackingTime should appear in the list of active MCP servers.
From that point, you can ask Gemini to log time, list projects, query your team's hours, or manage workspace data.

Go to Gemini

CONNECT WITH CHATGPT

Use TrackingTime as a Connector in ChatGPT

ChatGPT supports custom MCP servers through its Connectors feature. Once connected, your TrackingTime workspace is accessible directly from the chat. Available for Pro and Plus users with Developer Mode enabled.

01

Enable Developer Mode in ChatGPT

Open ChatGPT Settings, go to Connectors, select Advanced, and turn on Developer Mode. This unlocks the ability to connect custom MCP servers.

Go to ChatGPT
02

Add the TrackingTime MCP server

In Settings, go to Connectors and add a new custom MCP server. Enter the TrackingTime server URL and authenticate using your email and App Password.

MCP configuration
03

Start tracking from ChatGPT

Ask ChatGPT to log hours, list your projects, check your team's time data, or pull reports. Every response pulls live data from your TrackingTime workspace.

Track from GPT

CONNECT WITH YOUR CODING AGENT

Add TrackingTime MCP to Cursor, Windsurf, or VS Code

Cursor, Windsurf, and VS Code all support MCP through a configuration file. Once set up, your coding agent can log time and query workspace data without leaving the editor.

01

Generate an App Password

Log in to your TrackingTime workspace, go to Settings, and open Apps and Integrations.
Create a new App Password and copy it.


App settings
02

Add TrackingTime to your MCP configuration

Each coding agent has its own MCP settings panel or configuration file. In all cases, you add a TrackingTime server entry pointing to the MCP endpoint and include your App Password for authentication.

MCP configuration
03

Track time without leaving your editor

Once connected, your coding agent can log hours to projects and tasks, start and stop timers, and query workspace data from inside the development environment.


Go to Cursor

ANY OTHER MCP-COMPATIBLE CLIENT

Connect Any MCP Tool to TrackingTime

The TrackingTime MCP Server follows the open MCP standard. Any compatible client can connect using one of the two transport methods below.

01

Remote HTTP

Recommended for hosted and cloud-based tools. Connect directly to the TrackingTime MCP endpoint using HTTP transport.
No local setup or bridge required.



MCP configuration
02

Stdio via mcp-remote

For local clients that use subprocess-based MCP. Run the mcp-remote bridge via npx and point it to the TrackingTime endpoint.
Requires Node.js 18 or later.



MCP configuration
03

Authentication

All connections use your TrackingTime email and an App Password generated from Settings. Your workspace is auto-detected from your credentials. No account ID needed. App Passwords can be revoked at any time.

MCP configuration

TrackingTime MCP server

Connect your AI tools to accurate time data.

When your AI assistant can read and write time data directly, logging hours becomes part of the workflow. Every session in Claude, every task in Cursor, every conversation in Gemini can generate an accurate time entry without switching tools or filling in timesheets after the fact. The TrackingTime MCP Server makes this possible by giving your AI assistant direct access to your workspace — projects, tasks, users, and reports included. The data stays consistent, the records stay complete, and your team stays focused on the work.

Track time without switching tools

Log hours from Claude, Gemini, Cursor, or any MCP-compatible client using natural language. Tell your assistant what you worked on and which project it belongs to. The entry is created in your TrackingTime workspace instantly, with timestamp, project, and notes attached.

Query your team’s time data instantly

Ask your AI assistant how many hours the team tracked this week, which projects are running over budget, or what tasks are still open. Answers come live from your TrackingTime workspace.

Accurate timesheets for billing and compliance

Every time entry created through the MCP server is recorded in TrackingTime with timestamps, project assignments, and notes. Reports stay accurate and billing stays reliable.

Secure, hosted, and always available

The MCP server runs on TrackingTime infrastructure. Authentication uses App Passwords separate from your main login, so you can revoke access at any time. No local setup or maintenance required.

When your AI assistant has direct access to your workspace, time tracking becomes part of the conversation. Log hours, check reports, and manage projects from Claude, Gemini, or any MCP-compatible tool, with no tab switching required.

USAGE EXAMPLES

Talk to Your AI Assistant, We Do the Rest

Once your AI tool is connected via MCP, you can manage your entire TrackingTime workspace through natural language. Here are some examples of what you can ask.

Personal tracking and tasks

  • What tasks am I currently tracking time on, and how many hours have I logged this week?
  • What are the open tasks assigned to me across all projects, sorted by project?
  • Create a new task called ‘Client Review – Q2 Campaign’ under project X and start the timer
  • Log 2 hours on [Project Name] for this morning and add the note ‘Kickoff meeting with client’
  • What did I work on yesterday and how many hours did I log in total?
  • Stop my current timer and show me how much time I tracked today

Team and project visibility

  • Show me all tasks in the [Project Name] project that still have no time logged
  • Which team members haven’t logged any time in the past 3 days?
  • Show me all time entries submitted for approval that are still pending
  • List all projects that are currently active and show their task count
  • Add a comment to task [ID] saying ‘Hours reviewed and approved for invoicing’
  • Which projects have had no time logged in the last 7 days?

Client hours and billing

  • How many billable hours has the team logged this month, broken down by project?
  • What’s the total time logged against the [Client Name] retainer so far this month?
  • How many hours did we log for [Client Name] last month compared to this month — are we over or under the retainer?
  • Which clients have consumed more than 80% of their contracted hours this month?
  • Show me the billable vs. non-billable hour breakdown for each client this quarter

TIME TRACKING FOR AI-POWERED TEAMS

Automatic Time Tracking for AI Assistants

Connect your AI assistant to TrackingTime via MCP and keep track of every hour without leaving your workflow.

Start for free

14-day PRO trial included

Sign in with Google